System, Method and Computer Program Product for Region Specific Simultaneous Localization and Mapping

1. A computerized method for configuring navigation computer code for a Simultaneous Localization and Mapping (SLAM) based system, the method comprising: obtaining a three-dimensional (3D) model of a region of a first type of region; acquiring a first recording of visual data (VD) by at least one first VD collector traversing the region at first operational conditions and a second recording of VD by at least one second VD collector traversing the region at second operational conditions; applying SLAM to the first recording using a first plurality of sets of SLAM parameter values, thereby generating using each set of the first plurality of sets a corresponding first-recording trajectory assessing movement of the first VD collector within the region during the collection of the first VD recording; for each set of SLAM parameter values out of the first plurality of sets, generating a performance evaluation of a performance of the respective set of SLAM parameter values under the first operational conditions, based at least on accuracy of the respective first-recording trajectory with respect to the 3D model, thereby yielding a first plurality of performance evaluations pertaining to the first operational conditions; applying SLAM to the second recording using a second plurality of sets of SLAM parameter values, thereby generating using each set of the second plurality of sets a corresponding second-recording trajectory assessing movement of the second VD collector within the region during the collection of the second VD recording; for each set of SLAM parameter values out of the second plurality of sets, generating a performance evaluation of a performance of the respective set of SLAM parameter values under the second operational conditions, based at least on accuracy of the respective second-recording trajectory with respect to the 3D model, thereby yielding a second plurality of performance evaluations pertaining to the second operational conditions; based at least on the first plurality of performance evaluations and on the second plurality of performance evaluations, determining an operational set of SLAM parameter values for a SLAM-based system; and providing the operational set of SLAM parameter values to the SLAM-based system which is equipped with VD processor, for navigating within the first type of region by executing SLAM computer readable code by the VD processor using the operational set of SLAM parameter values.

2. The method according to claim 1, wherein a plurality of overlapping sets of SLAM parameter values is comprised in the first plurality of sets and in the second plurality of sets.

3. The method according to claim 1, wherein the operational set of SLAM parameter values outperforms a different operational set of SLAM parameter values in regions of the first type of region, wherein the different operational set of SLAM parameter values outperforms the operational set of SLAM parameter values in regions of a second type of regions, wherein the first type of region and the second type of region are different types of regions selected from a group of types of regions consisting of: indoor warehouse, indoor mall, outdoor area with roads.

4. The method according to claim 1, wherein the generating of the performance evaluations for each set out of a group of sets that comprises at least one set out of the first plurality of sets and at least one another set out of the second plurality of sets comprises: generating an accuracy evaluation and a resources consumption evaluation, wherein the determining comprises balancing between accuracy and resource consumption by the operational set of SLAM parameter values.

5. The method according to claim 1, wherein the first recording and the second recording are based on recorded VD collected by a sensor group comprising at least one VD sensor being installed on a navigable platform when the navigable platform traverses the region; wherein the 3D model is generated based on VD collected by at least one high-quality VD sensor installed on the navigable platform concurrently to the collection of the recorded VD.

6. The method according to claim 5, wherein the first VD recording is generated by downgrading VD collected by the at least one high-quality VD sensor.

7. The method according to claim 1, wherein the sets of the first plurality of sets and the sets of the second plurality of sets each comprise different values of: (a) maximal number of features to track, (b) degree of resolution downscaling for feature extraction, and (c) pose-graph loop closing parameters.

8. The method according to claim 1, wherein the operational set of SLAM parameter values and a reference set of SLAM parameter values differ only in that the operational set of SLAM parameter values provides less of an identifiable resource for performance of SLAM, wherein execution of SLAM based on the operational set of SLAM parameter values by the SLAM-based system provides better SLAM accuracy with respect to the 3D model than execution of SLAM based on the reference set of SLAM parameter values.

9. The method according to claim 1, wherein the determining comprises avoiding selection of sets of SLAM parameter values whose execution results in at least one of: (a) power consumption of the VD processor that exceeds a predetermined power consumption criteria, and (b) heating of the VD processor that exceeds a predetermined heating criteria.

10. A computerized method for configuring at least one navigation computer code for operation of SLAM-based systems in different types of regions, the method comprising: executing the method of claim 1 for providing a first operational set of SLAM parameter values to a first SLAM-based system which is equipped with a first VD processor, for navigating within a first type of region by executing SLAM computer readable code by the first VD processor using the first operational set of SLAM parameter values; and executing the method of claim 1 for a second type of region that is different than the first type of region, for providing a second operational set of SLAM parameter values to a second SLAM-based system which is equipped with a second VD processor, for navigating within the second type of region by executing SLAM computer readable code by the second VD processor using the second operational set of SLAM parameter values; wherein the first SLAM-based system and the second SLAM-based system are functionally equivalent, wherein the first VD processor and the second VD processor are functionally equivalent; wherein the first SLAM-based system outperforms the second SLAM-based system in regions of the first type of region, wherein the second SLAM-based system outperforms the first SLAM-based system in regions of the second type of region.

11. The method according to claim 10, wherein the second SLAM-based system is the first SLAM-based system that is operable to alternate between using the first operational set of SLAM parameter values when navigating the first type of region and using the second operational set of SLAM parameter values when navigating the second type of region.

12. The method according to claim 10, further comprising detecting deterioration in performance of SLAM using the first operational set of SLAM parameter values, and based on the detection triggering navigation using the second operational set of SLAM parameter values.

13. A computerized system for configuring navigation computer code for SLAM-based system, the system comprising: a processor; and non-tangible memory operable to store: a three-dimensional (3D) model of a region of a first type of region; a plurality of first-recording trajectories, each first-recording trajectory comprising an assessment of movement of a first visual data (VD) collector through the region during collection of a first VD recording by the first VD collector, wherein each first-recording trajectory is generated from the first VD recording by application of SLAM using a different corresponding set of SLAM parameter values out of a first plurality of sets; a plurality of second-recording trajectories, each second-recording trajectory comprising an assessment of movement of a second VD collector through the region during collection of a second VD recording by the second VD collector, wherein each second-recording trajectory is generated from the second VD recording by application of SLAM using a different corresponding set of SLAM parameter values out of a second plurality of sets; and computer program code; wherein the non-tangible memory and the computer program code are configured, with the at least one processor, to cause the computerized system to at least: for each set of SLAM parameter values out of the first plurality of sets, generate a performance evaluation of a performance of the respective set of SLAM parameter values under a first operational conditions, based at least on accuracy of the respective first-recording trajectory with respect to the 3D model, thereby yielding a first plurality of performance evaluations pertaining to the first operational conditions; for each set of SLAM parameter values out of the second plurality of sets, generate a performance evaluation of a performance of the respective set of SLAM parameter values under a second operational conditions, based at least on accuracy of the respective second-recording trajectory with respect to the 3D model, thereby yielding a second plurality of performance evaluations pertaining to the second operational conditions; based at least on the first plurality of performance evaluations and on the second plurality of performance evaluations, determine an operational set of SLAM parameter values for a SLAM-based system; and provide the operational set of SLAM parameter values to the SLAM-based system which is equipped with VD processor, for navigating within the first type of region by executing SLAM computer readable code by the VD processor using the operational set of SLAM parameter values.

14. The system according to claim 13, wherein a plurality of overlapping sets of SLAM parameter values is comprised in the first plurality of sets and in the second plurality of sets.

15. The system according to claim 13, wherein the operational set of SLAM parameter values outperforms a different operational set of SLAM parameter values in regions of the first type of region, wherein the different operational set of SLAM parameter values outperforms the operational set of SLAM parameter values in regions of a second type of regions, wherein the first type of region and the second type of region are different types of regions selected from a group of types of regions consisting of: indoor warehouse, indoor mall, outdoor area with roads.

16. The system according to claim 13, wherein the non-tangible memory and the computer program code are configured, with the at least one processor, to cause the computerized system to: generate as part of the performance evaluations an accuracy evaluation and a resources consumption evaluation for each set out of a group of sets that comprises (a) at least one set out of the first plurality of sets and (b) at least one another set out of the second plurality of sets, in the determining of the operational set of SLAM parameter values, balance between accuracy and resource consumption by the operational set of SLAM parameter values.

17. The system according to claim 13, wherein the first recording and the second recording are based on recorded VD collected by a sensor group comprising at least one VD sensor being installed on a navigable platform when the navigable platform traverses the region, wherein the 3D model is generated based on VD collected by at least one high-quality VD sensor installed on the navigable platform concurrently to the collection of the recorded VD.

18. The system according to claim 17, wherein the first VD recording is generated by downgrading VD collected by the at least one high-quality VD sensor.

19. The system according to claim 13, wherein the sets of the first plurality of sets and the sets of the second plurality of sets each comprise different values of: (a) maximal number of features to track, (b) degree of resolution downscaling for feature extraction, and (c) pose-graph loop closing parameters.

20. The system according to claim 13, wherein an only difference between the operational set of SLAM parameter values and a reference set of SLAM parameter values is that an allocation by the operational set of SLAM parameter values of a certain identifiable resource for performance of SLAM by the SLAM-based system is lower than an allocation by the reference set of SLAM parameter values of the certain identifiable resource for performance of SLAM by the SLAM based system, wherein execution of SLAM by the SLAM-based system based on the operational set of SLAM parameter values provides better SLAM accuracy with respect to the 3D model than execution of SLAM by the SLAM-based system based on the reference set of SLAM parameter values.

21. The system according to claim 13, wherein the non-tangible memory and the computer program code are configured, with the at least one processor, to cause the computerized system to avoid determining an operational set of SLAM parameter values whose execution results in at least one of: (a) power consumption of the VD processor that exceeds a predetermined power consumption criteria, and (b) heating of the VD processor that exceeds a predetermined heating criteria.

22. The system according to claim 13, wherein the non-tangible memory and the computer program code are further configured, with the at least one processor, to determine a plurality of operational sets of SLAM parameter values based on corresponding 3D models and VD recordings and on performance analysis of different sets of SLAM based parameters, and to provide the plurality of operational sets to at least one SLAM-based system for navigation within region of different types of regions, wherein each of the operational sets of parameters is usable for optimizing performance of SLAM by SLAM-based system of a class of SLAM-based systems for a different type of region out of the different types of regions.

23. A non-transitory computer-readable medium for configuring navigation computer code for a Simultaneous Localization and Mapping (SLAM) based system, comprising instructions stored thereon, that when executed on a processor, perform the steps of: obtaining a three-dimensional (3D) model of a region of a first type of region; acquiring a first recording of visual data (VD) by at least one first VD collector traversing the region at first operational conditions and a second recording of VD by at least one second VD collector traversing the region at second operational conditions; applying SLAM to the first recording using a first plurality of sets of SLAM parameter values, thereby generating using each set of the first plurality of sets a corresponding first-recording trajectory assessing movement of the first VD collector within the region during the collection of the first VD recording; for each set of SLAM parameter values out of the first plurality of sets, generating a performance evaluation of a performance of the respective set of SLAM parameter values under the first operational conditions, based at least on accuracy of the respective first-recording trajectory with respect to the 3D model, thereby yielding a first plurality of performance evaluations pertaining to the first operational conditions; applying SLAM to the second recording using a second plurality of sets of SLAM parameter values, thereby generating using each set of the second plurality of sets a corresponding second-recording trajectory assessing movement of the second VD collector within the region during the collection of the second VD recording; for each set of SLAM parameter values out of the second plurality of sets, generating a performance evaluation of a performance of the respective set of SLAM parameter values under the second operational conditions, based at least on accuracy of the respective second-recording trajectory with respect to the 3D model, thereby yielding a second plurality of performance evaluations pertaining to the second operational conditions; based at least on the first plurality of performance evaluations and on the second plurality of performance evaluations, determining an operational set of SLAM parameter values for a SLAM-based system; and providing the operational set of SLAM parameter values to the SLAM-based system which is equipped with VD processor, for navigating within the first type of region by executing SLAM computer readable code by the VD processor using the operational set of SLAM parameter values.

24. The non-transitory computer-readable medium according to claim 23, wherein a plurality of overlapping sets of SLAM parameter values is comprised in the first plurality of sets and in the second plurality of sets.

25. The non-transitory computer-readable medium according to claim 23, wherein the operational set of SLAM parameter values outperforms a different operational set of SLAM parameter values in regions of the first type of region, wherein the different operational set of SLAM parameter values outperforms the operational set of SLAM parameter values in regions of a second type of regions, wherein the first type of region and the second type of region are different types of regions selected from a group of types of regions consisting of: indoor warehouse, indoor mall, outdoor area with roads.

26. The non-transitory computer-readable medium according to claim 23, wherein the generating of the performance evaluations for each set out of a group of sets that comprises at least one set out of the first plurality of sets and at least one another set out of the second plurality of sets comprises: generating an accuracy evaluation and a resources consumption evaluation, wherein the determining comprises balancing between accuracy and resource consumption by the operational set of SLAM parameter values.

27. The non-transitory computer-readable medium according to claim 23, wherein the first recording and the second recording are based on recorded VD collected by a sensor group comprising at least one VD sensor being installed on a navigable platform when the navigable platform traverses the region, wherein the 3D model is generated based on VD collected by at least one high-quality VD sensor installed on the navigable platform concurrently to the collection of the recorded VD.

28. The non-transitory computer-readable medium according to claim 23, wherein the first VD recording is generated by downgrading VD collected by the at least one high-quality VD sensor.

29. The non-transitory computer-readable medium according to claim 23, wherein the sets of the first plurality of sets and the sets of the second plurality of sets each comprise different values of: (a) maximal number of features to track, (b) degree of resolution downscaling for feature extraction, and (c) pose-graph loop closing parameters.

30. The non-transitory computer-readable medium according to claim 23, wherein an only difference between the operational set of SLAM parameter values and a reference set of SLAM parameter values is that an allocation by the operational set of SLAM parameter values of a certain identifiable resource for performance of SLAM by the SLAM-based system is lower than an allocation by the reference set of SLAM parameter values of the certain identifiable resource for performance of SLAM by the SLAM based system, wherein execution of SLAM by the SLAM-based system based on the operational set of SLAM parameter values provides better SLAM accuracy with respect to the 3D model than execution of SLAM by the SLAM-based system based on the reference set of SLAM parameter values.

31. The non-transitory computer-readable medium according to claim 23, wherein the determining comprises avoiding selection of sets of SLAM parameter values whose execution results in at least one of: (a) power consumption of the VD processor that exceeds a predetermined power consumption criteria, and (b) heating of the VD processor that exceeds a predetermined heating criteria.

32. A non-transitory computer-readable medium for configuring at least one navigation computer code for operation of SLAM-based systems in different types of regions, comprising instructions stored thereon, that when executed on a processor, perform the steps of: executing operational instructions that comprise the instructions of claim 23, for providing a first operational set of SLAM parameter values to a first SLAM-based system which is equipped with a first VD processor, for navigating within a first type of region by executing SLAM computer readable code by the first VD processor using the first operational set of SLAM parameter values; and executing the operational instructions for a second type of region that is different than the first type of region, for providing a second operational set of SLAM parameter values to a second SLAM-based system which is equipped with a second VD processor, for navigating within the second type of region by executing SLAM computer readable code by the second VD processor using the second operational set of SLAM parameter values; wherein the first SLAM-based system and the second SLAM-based system are functionally equivalent, wherein the first VD processor and the second VD processor are functionally equivalent; wherein the first SLAM-based system outperforms the second SLAM-based system in regions of the first type of region, wherein the second SLAM-based system outperforms the first SLAM-based system in regions of the second type of region.

33. The non-transitory computer-readable medium according to claim 32, wherein the second SLAM-based system is the first SLAM-based system that is operable to alternate between using the first operational set of SLAM parameter values when navigating the first type of region and using the second operational set of SLAM parameter values when navigating the second type of region.

34. The non-transitory computer-readable medium according to claim 33, further comprising detecting deterioration in performance of SLAM using the first operational set of SLAM parameter values, and based on the detection triggering navigation using the second operational set of SLAM parameter values.